On Oct 30, 2007, at 1:11 PM, Carl Witty wrote: > On Tue, 2007-10-30 at 11:24 -0700, Robert Bradshaw wrote: >> I don't see this one--could you send me a link? >> >> I didn't know we used SAGE objects in any numpy classes. > > Numpy provides a type of matrices of arbitrary Python objects. > > http://groups.google.com/group/sage-support/browse_thread/thread/ > 96a5f3d336285146/a40540fd8a468ef3#a40540fd8a468ef3
Thanks. I really should subscribe to sage-support I guess, but I'm cross-posting to sage-devel 'cause I think it's an important issue. This is due to the inplace operator stuff using refcounts to determine if it's safe to mutate. The simple workaround is to not use numpy arrays of SAGE objects. Another question is why would one do so (i.e. what is lacking in the SAGE linear algebra types?) I think the benifits of inplace operators in terms of performance are to great to abandon, but am at loss to find a clean solution so that external extension classes don't mess them up... other than perhaps a global option that can be enabled/disabled. - Robert --~--~---------~--~----~------------~-------~--~----~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~----------~----~----~----~------~----~------~--~---
